def gexf_champ(years_bins,scores,champs,nb_label,dico_termes,sep_label,dist_type,distance_champ,niveau,seuil_net_champ):
legende_noeuds={}
legende_noeuds_id={}
for inter in range(len(years_bins)):
try:
os.mkdir(path_req + 'gexf')
except:
pass
nodes_chp = {}
level={}
dist_mat_chp_temp = {}
#on met d'abord des labels à tous les champs
for x,leschamps in champs.iteritems():
if inter==x:
for champ1 in leschamps:
if not leschamps.index(champ1) in nodes_chp:
etiquette,etiquette_id=labellize(leschamps.index(champ1),inter,scores,champs,nb_label,dico_termes,sep_label)
nodes_chp[leschamps.index(champ1)] = etiquette
level[leschamps.index(champ1)]=niveau
legende_noeuds[(inter,leschamps.index(champ1))]=etiquette
legende_noeuds_id[(inter,leschamps.index(champ1))]=etiquette_id
for x,dist in distance_champ.iteritems():
if inter==x[2]:
champ1 = x[0]
champ2 = x[1]
if not champ1 in nodes_chp:
etiquette,etiquette_id=labellize(champ1,inter,scores,champs,nb_label,dico_termes,sep_label)
nodes_chp[champ1] = etiquette
level[champ1]=niveau
legende_noeuds[(inter,champ1)]=etiquette
legende_noeuds_id[(inter,champ1)]=etiquette_id
if not champ2 in nodes_chp:
etiquette,etiquette_id=labellize(champ2,inter,scores,champs,nb_label,dico_termes,sep_label)
nodes_chp[champ2] = etiquette
legende_noeuds[(inter,champ2)]=etiquette
legende_noeuds_id[(inter,champ2)]=etiquette_id
if dist>seuil_net_champ and champ1!=champ2:
dist_mat_chp_temp[(champ1,champ2)]=dist
sortie = path_req + 'gexf/' + 'reseau_champ'+'_' +str(niveau)+'_'+ dist_type +'_'+str(years_bins[inter][0])+'-'+str(years_bins[inter][-1])+'.gexf'
for x in nodes_chp:
level[x]=niveau
gexf.gexf_builder(nodes_chp,dist_mat_chp_temp,sortie,level)
return legende_noeuds,legende_noeuds_id
dic = {}
else:
dic = dist_mat_temp_res[int(x[0])]
dic[int(x[1])]=float(y)
dist_mat_temp_res[int(x[0])] = dic
dist_mat_temp_res_10 = fonctions.seuiller(dist_mat_temp_res,10)
print 'on construit maintenant dist_mat pour chaque periode ' + str(inter)
for x,y in dist_mat_temp.iteritems():
dist_mat[(int(x[0]),int(x[1]),int(inter))]=y
for x,y in dist_mat_temp_res_10.iteritems():
for u in y:
dist_mat_10[(int(x),int(u[0]),int(inter))]=u[1]
level={}
for x in dico_termes:
level[x]=1
gexf.gexf_builder(dico_termes,dist_mat_temp,fichier_gexf,level)
fonctions.ecrire_dico(dico_termes,dico_termes,dico_termes,1)
print 'dicos ecrits'
#fonctions.dumpingin(p_cooccurrences,'p_cooccurrences'+name_date)
fonctions.dumpingin(dist_mat_10,'dist_mat_10'+name_date)
print 'on a dumpé distmat_10: top 10 de chaque terme'
fonctions.dumpingin(dist_mat,'dist_mat'+name_date)
print 'on a dumpé distmat'
try:
fonctions_bdd.detruire_table(name_bdd,'sem_weighted')
except: pass
#fonctions_bdd.creer_table_sem(name_bdd,'sem')
fonctions_bdd.creer_table_sem_periode_valuee(name_bdd,'sem_weighted')
lienssem_weighted=[]
def union_map(niv_max, seuil_net_champ):
print seuil_net_champ
for inter in range(len(years_bins)):
noeuds = {}
multi_level_edges = {}
map_edges = {}
level = {}
indexs = {}
index = 0
print " temps : " + str(inter)
if 1:
for niveau in range(niv_max):
niveau = niveau + 1
fichier_leg = open(
path_req
+ "legendes"
+ "/"
+ "legendes"
+ "niv_"
+ str(niveau)
+ "_"
+ dist_type
+ "_"
+ str(years_bins[inter][0])
+ "-"
+ str(years_bins[inter][-1])
+ ".txt",
"r",
)
lignes = fichier_leg.readlines()
for ligne in lignes:
index = index + 1
lignev = ligne[:-1].split("\t")
indexs[str(niveau) + "_" + str(lignev[0])] = index
noeuds[index] = lignev[1]
level[index] = str(niveau)
for niveau in range(niv_max):
niveau = niveau + 1
fichier_lex = open(
path_req
+ "lexique"
+ "/"
+ "lexique_one_step_"
+ "niv_"
+ str(niveau)
+ "_"
+ dist_type
+ "_"
+ str(years_bins[inter][0])
+ "-"
+ str(years_bins[inter][-1])
+ ".txt",
"r",
)
fichier_res = open(
path_req
+ "reseau"
+ "/"
+ "reseauCF_"
+ "niv_"
+ str(niveau)
+ "_"
+ dist_type
+ "_"
+ str(years_bins[inter][0])
+ "-"
+ str(years_bins[inter][-1])
+ ".txt",
"r",
)
lignes_res = fichier_res.readlines()
for ligne in lignes_res:
lignev = ligne[:-1].split("\t")
# print 'ici'
# print lignev
# print float(lignev[2])
if float(lignev[2]) > seuil_net_champ:
map_edges[(indexs[str(niveau) + "_" + lignev[0]], indexs[str(niveau) + "_" + lignev[1]])] = str(
lignev[2]
)
lignes_lex = fichier_lex.readlines()
print fichier_lex
if niveau > 1:
for ligne in lignes_lex:
lignev = ligne[:-1].split("\t")
for dessous in lignev[1].split(" "):
map_edges[
(
indexs[str(niveau) + "_" + str(lignev[0])],
indexs[str(niveau - 1) + "_" + str(dessous)],
)
] = "1"
map_edges[
(
indexs[str(niveau - 1) + "_" + str(dessous)],
indexs[str(niveau) + "_" + str(lignev[0])],
)
] = "1"
# except:
# pass
# print map_edges
sortie = (
path_req
+ "gexf/"
+ "reseau_multilevel_champ"
+ "_"
+ dist_type
+ "_"
+ str(years_bins[inter][0])
+ "-"
+ str(years_bins[inter][-1])
+ ".gexf"
)
gexf.gexf_builder(noeuds, map_edges, sortie, level)
